We propose to test the efficacy of the combination of naltrexone and fluoxetine versus fluoxetine alone in the treatment of patients with alcoholism and co-morbid major depression in a double-blind, placebo- controlled, randomized, parallel group trial. With nearly eight million affected individuals in the U.S., co-morbid alcoholism and major depressive disorder represent a significant public health problem. The presence of co-morbidity has a significant negative impact on treatment response and outcome, resulting in increased risk for suicide and increased rates of costly inpatient psychiatric care. Effective pharmacologic treatments addressing thee dual disorders are lacking. Only partial response has been obtained in studies evaluating anti- depressant monotherapy in depressed alcoholics. Our previous work with the SSRI fluoxetine has demonstrative the positive results published to date in severely depressed alcoholics. Our previous work with the SSRI fluoxetine has demonstrative the most positive results published to date in severely depressed alcoholics. The fluoxetine group in that study, however, displayed only a partial treatment response, with low abstinence rates and persistent depressive symptoms and alcohol abuse. However, our original and extended pilot work evaluating the usefulness of combined naltrexone and fluoxetine suggest a robust response, with a significant decrease in alcohol use and depressive symptoms. Our study of potential interactions between these two medications documents that naltrexone does not increase fluoxetine or norfluoxetine blood levels in most patients. Our proposed study will build on our previous work and established record both in conducting medication efficacy trials in this complex and high risk population, and in developing fundamental pharmacological methodologies necessary to investigate the proposed rug interaction studies. The timeliness of our proposed study is underscored by the high prevalence of this co-morbid condition and by the widely but untested use of the combined medication treatment in clinical practice. Thus, our study our will fill an important gap in our knowledge regarding the treatment of high risk clinical population. We hypothesize that combined fluoxetine and naltrexone treatment will offer enhanced treatment for alcoholics with co-morbid major depression. While the fluoxetine will target the depressive disorders in addition to the compulsive consumatory behavior related to alcoholism, the naltrexone will target the positive reinforcing effect and release risk related to pathological alcohol use. We request five years of support to achieve the following aims: 1) Examine the efficacy of naltrexone plus fluoxetine compared to fluoxetine and placebo in the treatment of patients with co- morbid DSM-IV alcohol dependence and unipolar major depression.; 2) Assess specific predictors of medication response; 3) Conduct a prospective assessment of the effect of persistent depressive symptoms on alcohol use. One hundred and six acutely depressed and actively drinking subjects will be randomized and prospectively followed during a 6 month double-blind study, and a 6-month post-treatment follow-up phase.